Applying Association Rule Mining to Explore Unsafe Behaviors in the Indonesian Construction Industry

The frequency of work accidents in construction projects is relatively high. One contributing factor to unsafe behavior by workers at sites. In Indonesia, this the first study investigate 2503 instances that occurred across Indonesian relation their attributes obtain insightful knowledge using association rule mining (ARM) method. Association was used explore database. As a result, two consolidated rules were obtained. most frequent behaviors putting tools and materials random places, not attaching safety lines provided moving ways accordance with procedures. These associated accident types falling, being struck or cut items, as well violations Manpower Transmigration Ministerial Regulation 01/1980, 09/2016. ARM results evaluated reliability evaluation method before validated experts. Hence, findings are reliable be guideline information for trainers prioritize related trainings inspectors when carrying out inspections on management performance can increase significantly.